La misura della luminosita in ATLAS nel 2010

La misura della luminosita` in ATLAS nel 2010 e 2011 Benedetto Giacobbe INFN Bologna

outline The ATLAS luminosity strategy Detectors and algorithms 2010 Luminosity calibration: Performed Vd. M scans and results (proton-proton) Achieved precision on luminosity determination Status of luminosity measurement at the beginning of 2011 Present precision Plans for improvements 2

The ATLAS strategy Several detectors and algorithms/detector available in ATLAS to: Keep systematics under control Adapt to beam conditions Check time stability Always have backup solution in case of detectorrelated problems On-line vs off-line: best on-line monitor can be different from best off-line one 3

2010: Detectors and algorithms (online) LUCID: official lumi-provider through most 2010 Event. OR and Event. AND calibrated in all scans Hit-counting calibrated in latest scans BCM: Event. OR calibrated in latest scans Event. AND calibrated in latest scans but not in note MBTS_Evt. AND: calibrated and used until bunchtrains started Long pulses limit to >150 ns bunch spacing ZDC_Evt. AND + A/C sides: available from HI runs Calibrated in HI Vd. M scans 4

LUCID: the main luminosity detector • Almost entire responsability on the Bologna group – – – Detector: project and construction Electronics and trigger Calibration, DCS, LUCID-DAQ, etc. . . Analysis and luminosity determination with LUCID MC simulation 5

2010: Offline methods MBTS and LAr timing methods Require A-C coincidence within 10 ns LAr timing not re-calibrated in latest scans (must investigate effect of significant pile-up) Primary-vertex Event counting Count BX with ≥ 1 vertex with ≥ 4 tracks of with p. T>150 Me. V + Pixel, SCT requirements Calibrated in all scans Charged-particle event counting ≥ 1 track with p. T>500 Me. V and |η|<0. 8 Used for ATLAS-CMS comparison (~2% agreement) Not yet recalibrated with latest scans 6

2010 systematic uncertainty on official luminosity (LUCID_Event. OR) Dramatic drop in systematic uncertainty: now 3. 4% ! Bunch currents still dominate Until Vd. M a limit of ~10% from machine parameters was guessed from past experiences ! 7

μ-dependence and time-stability • Last scans at 0<μ<1. 3: μ-dependence studied using data spanning large μ-range: – evaluated 0. 5% on σvis • Time stability: added conservative 0. 5% 8

Cross check with Z and W LZ/W = (NZ/W - Nbkg)/εσZ/W N, σ, ε and errors from Moriond Z/W note L is the official ATLAS luminosity lucid Agreement within systematic uncertainty on luminosity No μ-dependence observed Study performed by Alberto Mengarelli e Benedetto Giacobbe 9

What we did say/promise in February (ATLAS week 02/2011) before 2011 data taking 10

Scenario for 2011 (1/3) Scenario for LHC conditions not totally clear: Foreseen increase up to μ ~13 (could be even more) Combination of β*, emittance and bunch population 75 ns bunch spacing (now 50 ns) Challenging scenario Need to control μ-dependence in a unexplored range Some algorithms run toward saturation (no sensistivity to L variations) Lot of experience gained in 2010 but new problems related to μ increase are to be faced 11

summary for 2011 Many improvements in the luminosity infrastructure in order to face 2011 challenge LUCID/BCM and ZDC (Bologna has taken responsability on it) Monitoring of the detectors crucial to ensure the present L precision Time-stability Consistency among detectors and algorithms μ-dependence There’s a lot to do to ensure the 2010 precision ! It is not garanteed at least at the beginning 12

What we did find at the beginning of 2011 data taking 13

What happened in 2010 -2011 shutdown LUCID: No interventions BUT residual magnetic field changed inside shielding => effect on calibration expected BCM: upgrade of Front-end electronics => effect on calibration expected 14

15

16

17

18 d correlate

19

2011 “preliminary-final” comments We promised we would not have done better than in 2010 at the beginning Promise kept. . . BUT We are not far away from 2010 given the complete new scenario from LHC We have a lot of handles to improve: Cross-consistency among methods (Vd. M calibration) μ-dependence + afterglow background Before 2010 we would have “leccati i gomiti” for a 4. 5% uncertainty but we understand that “la fame vien mangiando”. . . 20

Backup slides

Documentation Updated Luminosity Determination in pp Collisions at sqrt(s)=7 Te. V using the ATLAS detector ATLAS-COM-CONF-2011 -034 http: //cdsweb. cern. ch/record/1330358 Lucid COM Note: http: //cdsweb. cern. ch/record/1325334/ Offline COM Note: http: //cdsweb. cern. ch/record/1326897/ BCM COM Note: http: //cdsweb. cern. ch/record/1328923/ Bunch currents: http: //cdsweb. cern. ch/record/1325370/ 22

Online algorithms Algorithm Early 2010 Final 2010 2011 LUCID_Evt. OR calibrated μ<13 LUCID_Evt. AND calibrated μ<15 LUCID_Hit. OR calibrated yes LUCID_OR-A/C Not existing yes BCM_Evt. OR not calibrated yes BCM_Evt. AND not calibrated yes BCM_x-OR-A/C not calibrated yes MBTS_Evt. OR calibrated No (afterglow) no MBTS_Evt. AND calibrated No (afterglow) no ZDC_Evt. AND no Heavy Ions yes ZDC_A / ZDC_C no Heavy Ions yes 2011 foreseen availability independent of calibration 23